Five AI Moves Subscribe free

Safe to paste? Check your text before you give it to AI

Paste an email thread, a document or a spreadsheet export. See the personal data, passwords and keys hiding in it, and get a clean copy that's safe to share with an AI tool.

Free. No signup. Your text never leaves your browser and is not saved.

The check runs in your browser. Nothing you paste is sent or saved, and the page's security settings block background data requests.

What this checker looks for

  • Secrets: passwords, API keys and tokens (OpenAI, Anthropic, AWS, GitHub, Google, Slack, Stripe), private keys, and secrets hidden in links.
  • Money and ID numbers: payment card numbers (checked with the card checksum), bank account numbers in IBAN format (checked with the IBAN checksum), US Social Security numbers and UK National Insurance numbers.
  • Contact details: email addresses and phone numbers.
  • Network details: IP addresses.

Each item is replaced with a label such as [EMAIL 1] or [CARD NUMBER 1]. The same value always gets the same label, so the AI can still follow a conversation between [EMAIL 1] and [EMAIL 2].

What you should never paste into ChatGPT or any AI tool

  1. Passwords, API keys and access tokens.
  2. Payment card numbers and bank account details.
  3. ID numbers, such as social security, national insurance or passport numbers.
  4. Customer personal data, unless your company uses a business plan it has approved for this.
  5. Documents marked confidential, contracts under NDA and unreleased financial figures.

Is it safe to paste customer data into ChatGPT?

It depends on the plan and its settings. OpenAI says it does not train on data from business products such as ChatGPT Business, ChatGPT Enterprise and the API by default; on personal accounts, the "Improve the model for everyone" setting decides whether your chats help train models. Anthropic says it does not train on Claude for Work data; on personal Claude plans, the Model Improvement setting decides. Google says a subset of Gemini app chats is read by human reviewers, and asks you not to enter confidential information you wouldn't want a reviewer to see.

Whatever the plan, removing details the AI doesn't need is the safest default. Check the current terms at the source: OpenAI, Anthropic (business), Anthropic (personal plans), Google Gemini.

What the checker can't catch

It finds data with a recognisable format. It can't recognise names, street addresses, health details, salaries or confidential business information, because those look like ordinary text. Give anything sensitive a quick read before you paste it.

Common questions

Does this tool store or send my text?

No. The check runs entirely in your browser and nothing is saved. The page has no code that sends your text, and its security settings block background data requests. Close the tab and the text is gone.

Why replace data with labels instead of deleting it?

Labels keep the text readable. The AI can still write "Reply to [EMAIL 1] and confirm the refund to [CARD NUMBER 1]", and you put the real details back before you send anything.

More free tools

  • Your First 5 AI Moves: Answer four questions and get five AI moves picked for you, from first steps to a power move.
  • AI ROI Calculator: See what your AI tools cost, what the time they save is worth, and where you pay twice.

Five AI Moves

Every week, 5 small AI moves anyone can try in 10 minutes, each with a clear next step and the original source. Free.

Subscribe free